Measuring the Effects of the September 11 Attack on New York City

Jason Bram, James Orr, and Carol Rapaport Measuring the Effects of the September 11 Attack on New York City • The total cost of the September 11 attack on the World Trade Center—comprising earnings losses, property damage, and the cleanup and restoration of the site—is estimated to bebetween $33 billion and $36 billion through June 2002. • The earnings losses consist of $7.8 billion in deceased workers’ prospective lifetime earnings and $3.6 billion to $6.4 billion in reduced wage and salary income in city industries affected by the attack. • The cost of cleaning up the site, replacing the destroyed World Trade Center buildings, and repairing damaged buildings and infrastructure is expected to reach $21.6 billion. • Although the loss of life and disruption of activity temporarily reduced New York City’s productive capacity, the attack’s effects on employment and consumer confidence had largely run their course by mid-2002. Jason Bram is an economist, James Orr a research officer, and Carol Rapaport an economist at the Federal Reserve Bank of New York. he attack on the World Trade Center on September 11, 2001, traumatized New York City and the nation. Almost 3,000 lives were lost, and more than 30 million square feet of office space in Lower Manhattan was damaged or destroyed. The loss of workers, physical capital, and infrastructure reduced the productive potential of the city’s economy and disrupted the lives of hundreds of thousands of people. Damage to the transportation and communications infrastructure depressed economic activity for a numberof months, especially in Lower Manhattan. This article evaluates the short-term economic conse-quences of the attack on Manhattan and the four other boroughs that make up New York City. We begin with the deepest loss— that of human lives. We then look at the effects of the attack on the inputs to the production process: labor and capital. The attack led to an idling and underutilization oflabor not only in the World Trade Center area, but also in other parts of the city. (Views of New York City and Lower Manhattan are provided in Appendix A.) Our analysis of labor focuses on aggregate city employment as well as on industry effects and factors that impact employee productivity, including health and confidence.1 The analysis of capital covers the destruction of commercial space and infrastructure. We also discuss the effects of the attack on the markets for office space, home construction, and home sales. Finally, we examine how the attack affected the city’s most economically vulnerable residents. The authors thank Simon Potter and Robert Rich for valuable input and insights, as well as two anonymous referees. They also acknowledge the excellent research assistance of Silvia Ellis, David Lagakos, and Alisdair McKay. The views expressed are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. FRBNY Economic Policy Review / November 2002 5 This framework is an economic one, incorporating quality-of-life issues. To pursue our analysis, we have restricted ourselves to the labor and capital markets. In theory, it should be possible to evaluate output and income losses directly. In practice, however, such an evaluation is unworkable because official tabulations of gross NewYork Cityproduct do not exist and income figures are reported with a considerable lag.2 Thus, output effects must be inferred from the behavior of the labor and capital markets. Whenever possible, we separate the effects of the attack fromthe effects of the businesscycle (although we do not attempt to isolate the effects of the fall 2001 anthrax scares from the effects of the attack). Unless otherwise noted, the data presented here cover the period through June 2002, the end of the recovery process at the World Trade Center site. We conclude that the attack disrupted New York City’s economy in many ways. Although it is difficult to put a dollar value on lives lost, it is also inappropriate to omit loss of life from an estimate of the damage sustained. Our intention is to present as complete a picture of the attack’s effects as possible. Accordingly, we estimate that the aggregate present value of lost lifetime earnings for these workers is about $7.8 billion. In addition, in the ninemonths following the attack, lost jobsand a reduction in the number of hours worked translated into an estimated shortfall in aggregate earnings of $3.6 billion to $6.4 billion. The cost of replacing the destroyed and damaged physical capital and infrastructure is estimated at$21.6 billion. Finally, the sum of these labor and capital losses yields an estimated total loss through June 2002 of between $33 billion and $36 billion.3 We estimate workers who died in the attack earned, on average, $127,000 a year. This estimate is based on the average income in 2000 for all workers in Manhattan and all workers in the finance and insurance sectors in Manhattan. The average annual income for workers in the finance and insurance sectors—where about half of the deceased workers had been employed—is estimated to be $197,275 in 2002. The average annual income of all workers in Manhattan, excluding the two sectors, is estimated at $57,000.6 We use the average age of the workers killed in the attack, forty, and assume that they had twenty-two more years left to work until retirement. The average income of these workers is assumed to grow at the rate of inflation, which is assumed to equal the average discount rate. Under these assumptions, the current value of the aggregate earnings loss reaches about $7.8 billion, or an average of $2.8 million per worker.7 Although private insurance is expected to covera portion of these losses, it is not likely that all of the workers had taken out private life-insurance policies. The earnings losses sustainedby the workers’ families will be partially covered by various charitable funds. In addition, the families of all World Trade Center attack victims will be eligible to receive compensation under the federal VictimCompensationFund.8 Although these various payments will partially offset losses to families and individuals, they do not reduce the overall cost of the attack because those payments represent costs to other parties,such as the government and insurance companies. Employment Disruptions Loss of Life The death of almost 3,000 people in the attack was a loss to New York City and to the nation. This number includes those who worked in the two World Trade Center towers, the firefighters and police personnel who responded to the attack, and the tourists and other visitors who were in the World Trade Center complex that morning. The method we use to value loss of life is based on the concept of “lifetime-earnings loss.” This method estimates individual economic losses by adding up a worker’s pretax annual income from the year of death to the year that he or she had expected to retire.4 For those who died in the attack, the estimated earnings loss is calculated by multiplying the average expected level of annual earnings by the average number of years left to work before retirement.5 In addition to the loss oflives, theattackon the WorldTrade Center had a dramatic disruptive effect on employmentin New York City. The number of private-sector workers started to decline at the beginningof 2001 because of national and local business cycles.Thelevel of employmentbottomedout in March 2002 and edged upduring thesecond quarter of the year (Chart1).From the peak inemployment in December 2000 to the troughin March 2002, the number ofpeople working in New YorkCity’s private sectorfell by 147,000, or 4.6percent. (By comparison, the number of private-sector jobs lost duringthe 1989-92 recessionwas 344,000, or11.4percent.) Inthissection, we estimate thenumber of jobs lost because of the attack separatelyfrom those jobslost becauseof the business cycle. More than one-third of the net job losses in the recent downturn—specifically, 55,000 of the 147,000—occurred between January and September 2001. However, the sharpest 6 Measuring the Effects of the September 11 Attack drop was in October 2001: a record 51,000 private-sector jobs were lost in that month alone. The remaining 41,000 job losses of the peak-to-trough decline occurred between October 2001 and March 2002. However, in the following months of April, Chart 2 Path of New York City Private-Sector Employment Thousands 3,250 May, and June, the number of private-sector jobsrose by a total 3,200 of 10,000, or 0.4 percent. To gauge how much of the fall in the number of jobs can be 3,150 attributed to the attack, we use a standard dynamic forecasting 3,100 model to estimate what the path of New York’s employment would have been in the absence of an attack (Appendix B). The 3,050 difference between the actual path of employment and this 3,000 estimated path can be interpreted as the marginal effect of the 2,950 attackon employmentin the cityat monthlyintervals. Usingthis 2000 High-impact scenario Low-impact scenario Actual 2001 2002 technique and two alternative sets of assumptions (high-impact scenario and low-impact scenario), we estimate that in October 2001, the number of private-sector jobs in the city was about 38,000 to 46,000 lower than it would have been otherwise. In February, this range moved to as high as 49,000 to 71,000, then eased to between 28,000 and 55,000 by June 2002 (Chart 2). Data on weekly initial claims for unemployment insurance seem to confirm the pattern seen in payroll employment: the attack’seffects on employment were substantial in Octoberand November of 2001, but had largely run their course by early 2002 (Chart 3). Prior to September 11, weekly claims in New York City had been fluctuating in the 7,000 to 9,000 range—or about 1,000 to 3,000 higher than a year earlier, reflecting a general weakening in the economy. The weekly volume of claims more than doubled in the second half ofSeptember, and was running 10,000 to 12,000higher than a year earlier, but then Chart 1 Private-Sector Employment in New York City Thousands 3,300 3,200 3,100 3,000 2,900 2,800 2,700 Sources: U.S. Department of Labor, Bureau of Labor Statistics; Federal Reserve Bank of New York. Note: See Appendix B for methodology and a full explanation. retreated steadily for four months, returning to approximate pre-attack levels by late February 2002. Aside from a brief spike in late March and early April—largely attributable to filings for extended benefits—the number of jobless claims was relatively steady throughout the first half of 2002. These employment disruptions varied across the city’s boroughs and neighborhoods, and across industries (Box 1). The most pronounced impact was concentrated in the blocks surrounding the World Trade Center, where numerous businesses, offices, and retail shops were either destroyed or Chart 3 New York City Initial Jobless Claims Net change in thousands from a year earlier 12,000 8,000 4,000 0 September 11 -4,000 2,600 1988 89 90 91 92 93 94 95 96 97 98 99 00 01 02 Source: New York State Department of Labor. -8,000 Aug Sep Oct Nov Dec Jan 2001 Feb Mar Apr May Jun 2002 Note: The shading indicates the post-September 11 period. Source: New York State Department of Labor. FRBNY Economic Policy Review / November 2002 7 Box 1 Employment Disruptions by Industry The dynamic forecasting model suggests that most of the attack’s net impact on employment levels occurred in October 2001. Here, we take a closer lookat what appear to be the mostdirectlyaffected industries: financial services, restaurants, hotels, and air trans-portation. Together, these industries accounted for 42,000 of October’s 51,000 drop in private-sector employment. In subsequent months, although the estimated effect on overall employment was relatively modest, some industries registered further losses while others rebounded (see chart below). To get a better understanding of the attack’s effectsover time, it is helpful to examine these industries and their performance. Because swings in Financial Services Jobs in New York City and New Jersey Seasonally Adjusted Level Thousands 190 185 180 175 Thousands New Jersey 60 Scale 55 50 45 employment after September 11 are far larger than any preexisting trends within these industries, we assume that changes in employment after that date are mainly attributable to the attack. Thefinancial services industry appears to have been the most directly affected sector by far. In New York City, the number of 170 165 2000 2001 New York City Scale 2002 40 35 jobs in the securities industry fell by 12,000, or 7 percent, in October 2001, and by an additional 6,000 from October 2001 to June 2002. In addition, the banking industry saw a net job loss of 8,000, or 8 percent, in October and lost another 1,000 jobs through June 2002. Net job losses in these keyfinancial industries totaled 20,000 in October and another 7,000 through June 2002. Because some of the loss reflected a relocation of operations to nearby suburbs—mostly northern New Jersey—this figure overstates the net impact on the metropolitan area overall (see chart at right). The restaurant industry also sustained steep job losses immediately following the attack. For the city overall, the number of jobsat bars andrestaurants—whichwas imperceptibly affected at the national level—fell by an estimated 9,000 (6 percent) in October, but rebounded fully by December and held steadyup to June 2002. However, these are net changesand donot capture the geographical distribution of employment in this industry. Thus, Employment in Selected New York City Industries Net change in thousands from a year earlier 10 Restaurants and bars 0 Hotels Sources: New York State Department of Labor; New Jersey Department of Labor; Federal Reserve Bank of New York. Note: The shading indicates the post-September 11 period. it is not clear if restaurant employment in the areas closest to the World Trade Center—the Financial District, Tribeca, and Chinatown—has fully rebounded to pre-attack levels. The hotel industry lost an estimated 6,000 jobs, or 15 percent, citywide between September 2001 andMarch 2002. Thisreflected the drop-off in tourism, although 5,000 of those jobs were lost in October alone. In April 2002, the number of hotel jobs rose markedly by an estimated 4,000 and held steady in May at about 5 percent below pre-attack levels. Nationally, hotel industry employment has fallen by a more modest 4 percent since September 2001, but has yet to show any sign of bottoming out. The steep decline in the number of people traveling also ledto job losses in areas away from the World Trade Center site—in particular, at John F. Kennedy International Airport and LaGuardia Airport, both in the borough of Queens. The number of jobs in the city’s air transportation industry fell by about 11,000, or 20 percent. Almost all of this decline occurred in October and November 2001, and there has been no sign of a rebound. Nationally, the number of jobs in this industry fell by 10 percent, with losses spread over the fourth quarter of 2001. Although other industries, such as business services, apparel -10 -20 Securities and banking -30 2000 2001 Air transportation 2002 manufacturing, printing, and publishing, were also presumably affected, largely because of their strong concentration in Lower Manhattan, there is no indication of any significant shift in employment trends following September 11. However, it should be noted that many business owners and workers who did not lose their jobs evidently suffered income losses because of the disruptions in the weeks and months immediately following the Source: New York State Department of Labor. Note: The shading indicates the post-September 11 period. attack. This is of particular concern in the restaurant and apparel industries, where workers’ pay depends on business volume. 8 Measuring the Effects of the September 11 Attack badly damaged. Substantial employment effects were also felt in the whole of Lower Manhattan (south of Canal Street [Appendix A]), where transportation access was curtailed and the volume of customer traffic fell precipitously. However, because of the drop-off in tourism—as well as possible multiplier effects from the loss of finance jobs—businesses throughout the city suffered because of the attack. For example, John F. Kennedy International Airport and LaGuardia Airport (both in the borough ofQueens)saw, asdid related businesses, a sharp decline in employment in the fourth quarter of 2001. It is less clear whether the job losses were across all income levels. One might hypothesize that low-skilled, low-paid workers were more at riskoflosing their jobs; laboreconomists We estimate that the attack led to a shortfall in wage and salary earnings of $3.6 billion to $6.4 billion as of June 2002. This estimate mainly reflects attack-related job losses but also includes the reduction in the number of hours worked. generally maintain that the workers with the least job-specific skills are the first to be laid off in times of economic stress. Indeed, many of the workers in the hard-hit restaurant and retail sectors are relatively low-paid.9 To test the hypothesis that the city’s low-wage workforce faced a higher incidence of attack-related job loss than high-wage workers, we compare three industries where most employees are relatively well paid with three other industries where most employees are relatively poorly paid.10 Both the high- and low-wage industries experienced a range of employment declines. Employeesin the (low-wage)hotel and (high-wage) brokerage industries were especially affected. However, those in the (low-wage) general merchandise store and (high-wage) legal industries maintained previous employment trends. This example, although limited, does not support the hypothesis that the September 11 attack caused disproportionate job losses in low-wage industries. The attack also led to a reduction in the number of hours worked. A recent study of the effects of the attack on workersin Chinatown indicates substantial short-term disruptions in the restaurant and garment industries.11 Restaurants faced particularly severe declines in business volume in the weeks following the attack. These declines appear to have affected the number of hours worked as well as the number of jobs available. The garment industry also reported substantial declines in the number of hours worked (see Asian American Federation of New York [2002]). On the basis of this analysis, we estimate that the attack led to a shortfall in wage and salary earnings of $3.6 billion to $6.4 billion as of June 2002. This estimate mainly reflects attack-related job losses, but also includes the reduction in the number of hours worked (Box 2). Furthermore, worker productivity may have been lowered by changes in personal habits, health, and confidence. Vlahov et al. (2002) report the results from phone interviews with 988 adult Manhattan residents living south of 110th Street five to eight weeks after the attack. About 30 percent of the sample reported an increased use of cigarettes, alcohol, and/or marijuana. The same residents who increased their use of cigarettes and/or alcohol were also found to be more likely to have post-traumatic stress disorder (PTSD) and major depression. In a related study, Galea et al. (2002) report that about 7 percent of the phone sample reported psychological symptoms consistent with current PTSD and almost 10 percent reported symptoms consistent with depression. These percentages are about twice baseline values. In addition, the New York City Department of Health and the Centers for Disease Control performed a door-to-door survey of 414 individuals living in the Battery Park City residential complex (next to the World Trade Center site) and two other downtown areas most directly affected by the attack (Centers for Disease Control and Prevention 2002). As of October 2001, almost 40 percent of the sample showed PTSD symptoms. Moreover, about 50 percent were still experiencing symptoms consistent with smoke inhalation from the still-burning fires. Surveys of consumer confidence can also help shed light on the attack’s psychological effect on behavior. The widely cited [Consumer]confidence fell fairly sharply in September 2001, recovered somewhat in October, and then rebounded to above pre-attack levels in November. Conference Board survey is only available by census region (that is, New Jersey, New York, and Pennsylvania combined), but since 1997, Siena College in Loudonville, New York, has conducted a parallel monthly survey of New York State residents in which consumer confidence is reported separately for the New York City metropolitan area. According to the FRBNY Economic Policy Review / November 2002 9 ... - tailieumienphi.vn